Azoma has launched the Agentic Merchant Protocol (AMP), a framework that lets brands push centralized product data to AI agents, online marketplaces, and product listing platforms from a single source.
The four-year-old startup built AMP specifically for high-volume retailers in consumer packaged goods and fast-moving consumer goods — think grocery brands, fashion labels, and electronics manufacturers. It does not currently support NFTs, SaaS, or financial services sectors, according to the announcement.
The commercial logic behind the product is straightforward. Morgan Stanley research estimates that 10–20% of total U.S. commerce spend could be handled by AI agents by 2030 — a range of $190 billion to $385 billion. Brands that cannot feed structured, accurate data to those agents risk losing the sale before a human ever sees a product page.
The Problem AMP Is Solving
Today, merchants manually enter product information — SKUs, materials, specifications — across each marketplace separately. When an AI agent then researches a product on a customer’s behalf, it often pulls from unverified sources like Reddit threads or outdated affiliate pages. The brand’s intended messaging gets lost in what Azoma calls a “black box.”
Existing protocols such as OpenAI‘s ACP and Google‘s UCP handle technical discovery and payment handshakes but offer little control over how a brand is represented during an AI agent’s reasoning process.
AMP positions itself above those systems as a “system of record.” Brands load their product intelligence — including legal guardrails and brand guidelines — into Azoma‘s platform in a machine-native format, then distribute it everywhere simultaneously.
“For decades, marketplaces like Amazon and Walmart acted as gatekeepers by controlling product detail pages, rankings, and distribution,” said Max Sinclair, CEO of Azoma, in a press release. “In an agentic world, those fixed pages no longer exist.”
Early Adopters and What AMP Delivers
L’Oréal, Unilever, Mars, Beiersdorf, and Reckitt have already adopted the protocol. For organizations that have spent decades building brand equity, ceding control of product representation to an AI system drawing on unverified data is not an acceptable default.
“The fact that businesses like L’Oréal, Unilever, Mars and Beiersdorf have moved so quickly to adopt AMP tells you everything about the urgency they feel,” Sinclair said.
The protocol includes canonical machine-native catalogues — data structures enriched with persona-level signaling and built for LLM ingestion — alongside programmatic open web distribution to ensure product data reaches AI surfaces accurately and consistently.
The official launch is scheduled for March 12 in London.
Photo by Markus Spiske on Unsplash
This article is a curated summary based on third-party sources. Source: Read the original article
